Efficient Medical Diagnosis Hybrid System based on RF-DNN Mixed Model for Skin Diseases Classification

Abstract

Accurate and efficient diagnosis is essential for effective treatment and management of these diseases. The current diagnostic methods rely mostly on visual inspection by dermatologists, which can be subjective and time-consuming. Therefore, there is a need for an automated and accurate system for skin disease diagnosis. A hybrid system has the potential to improve the diagnostic accuracy and efficiency of skin disease classification. The proposed research presents an efficient medical diagnosis hybrid system that combines a Random Forest model and a Deep Neural Network for the classification of skin diseases. The system aims to improve diagnostic accuracy and efficiency by utilizing the strengths of both algorithms, such as their ability to handle large datasets, provide fast and accurate predictions, and analyze images of the patient's skin. The system is composed of two parts, a Random Forest classifier and a DNN classifier, and is evaluated on a dataset of skin disease images, achieving an accuracy of 96.8%. To optimize efficiency, the DNN is trained on a subset of data where the Random Forest model is less confident, and the system is able to identify important features for skin disease classification. The benefits of this hybrid system are clear, including increased accuracy and reliability, reduced time and cost associated with diagnosis, and its potential for continued use in the future.